Remote sensing of high resolution urban impervious surfaces - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 228

17
Remote sensing of high resolution
urban impervious surfaces
Changshan Wu and Fei Yuan
Numerous approaches have been developed to quantify the distribution of impervious surfaces through remote
sensing technologies in recent years. These methods can be divided into two groups: pixel-based models and
object-based models. Pixel-based models utilize each individual pixel as the basic unit of analysis, while
object-oriented models employ pre-identified spatial objects as the basic unit. While most of these methods have
been applied to medium resolution remote sensing data (e.g., Landsat TM), high-resolution impervious surface
information has also been generated. This chapter reviews major pixel-based and object-based techniques for
impervious surface estimation, and compares these two groups of models through case studies in Grafton, WI and
Mankato, MN, USA. Results indicate that for the pixel-based model, an integrated spectral mixture analysis and
regression tree model outperforms the individual models, and the object-based model performs well if the
parameters of image segmentation are defined appropriately.